The invention relates generally to screening methods involving use of membrane potential indicator dyes for identifying antimicrobial agents, including antifungal and antibacterial compounds.
Fungi are not only important human and animal pathogens, but they are also among the most common causes of plant disease. Fungal infections (mycoses) are becoming a major concern for a number of reasons, including the limited number of antifungal agents available, the increasing incidence of species resistant to known antifungal agents, and the growing population of immunocompromised patients at risk for opportunistic fungal infections, such as organ transplant patients, cancer patients undergoing chemotherapy, burn patients, AIDS patients, or patients with diabetic ketoacidosis. The incidence of systemic fungal infections increased 600% in teaching hospitals and 220% in non-teaching hospitals during the 1980""s. The most common clinical isolate is Candida albicans (comprising about 19% of all isolates). In one study, nearly 40%of all deaths from hospital-acquired infections were due to fungi. [Sternberg, Science, 266:1632-1634 (1994).]
Known antifungal agents include polyene derivatives, such as amphotericin B (including lipid or liposomal formulations thereof) and the structurally related compounds nystatin and pimaricin; flucytosine (5-fluorocytosine); azole derivatives (including ketoconazole, clotrimazole, miconazole, econazole, butoconazole, oxiconazole, sulconazole, tioconazole, terconazole, fluconazole, itraconazole, voriconazole [Pfizer] and SCH56592 [Schering-Plough]); allylamines-thiocarbamates (including toinaftate, naftifine and terbinafine); griseofulvin; ciclopirox; haloprogin; echinocandins (including MK-0991 [Merck]); and nikkomycins. Recently discovered as antifungal agents are a class of products related to bactericidal/permeability-increasing protein (BPI), described in U.S. Pat. Nos. 5,627,153, 5,858,974, 5,652,332, 5,763,567 and 5,733,872, the disclosures of all of which are incorporated herein by reference.
Resistance of bacteria and other pathogenic organisms to antimicrobial agents is an increasingly troublesome problem. The accelerating development of antibiotic-resistant bacteria, intensified by the widespread use of antibiotics in farm animals and overprescription of antibiotics by physicians, has been accompanied by declining research into new antibiotics with different modes of action. [Science, 264: 360-374 (1994).]
Gram-positive bacteria have a typical lipid bilayer cytoplasmic membrane surrounded by a rigid cell wall that gives the organisms their characteristic shape, differentiates them from eukaryotic cells, and allows them to survive in osmotically unfavorable environments. This cell wall is composed mainly of peptidoglycan, a polymer of N-acetylglucosamine and N-acetylmuramic acid. In addition, the cell walls of gram-positive bacteria contain teichoic acids which are anchored to the cytoplasmic membrane through lipid tails, giving rise to lipoteichoic acids. The various substituents on teichoic acids are often responsible for the biologic and immunologic properties associated with disease due to pathogenic gram-positive bacteria. Most pathogenic gram-positive bacteria have additional extracellular structures, including surface polysaccharides, capsular polysaccharides, surface proteins and polypeptide capsules.
Gram-negative bacteria also have a cytoplasmic membrane and a peptidoglycan layer similar to but reduced from that found in gram-positive organisms. However, gram-negative bacteria have an additional outer membrane that is covalently linked to the tetrapeptides of the peptidoglycan layer by a lipoprotein; this protein also contains a special lipid substituent on the terminal cysteine that embeds the lipoprotein in the outer membrane. The outer layer of the outer membrane contains the lipopolysaccharide (LPS) constituent.
Antibacterial agents are generally directed against targets not present in mammalian cells. One major difference between bacterial and mammalian cells is the presence in bacteria of a rigid wall external to the cell membrane. Thus, chemotherapeutic agents directed at any stage of the synthesis, export, assembly, or cross-linking of peptidoglycan lead to inhibition of bacterial cell growth and, in most cases, to cell death. These agents include bacitricin, the glycopeptides (vancomycin and teichoplanin), xcex2-lactam antibiotics (penicilins, cephalosporins, carbapenems, and monobactams). Virtually all the antibiotics that inhibit bacterial cell wall synthesis are bactericidal. However, much of the loss of cell wall integrity following treatment with cell wall-active agents is due to the bacteria""s own cell wall-remodeling enzymes (autolysins) that cleave peptidoglycan bonds in the normal course of cell growth. In the presence of antibacterial agents that inhibit cell wall growth, autolysis proceeds without normal cell wall repair; weakness and eventually cellular lysis occur. There are also antibacterial agents that do not affect cell wall synthesis but instead are believed to alter cell membrane permeability, such as the polymyxins (polymyxin B and colistin, or polymyxin E) and gramicidin A.
Another group of antibacterial agents are those that inhibit protein synthesis; most of these interact with the bacterial ribosome. The difference between the composition of bacterial and mammalian ribosomes gives these compounds their selectivity. These agents include the aminoglycosides (e.g., gentamicin, kanamycin, tobramycin, streptomycin, netilmicin, neomycin, and amikacin), the macrolides (e.g., erythromycin, clarithromycin, and azithromycin), the lincosamides (e.g., clindamycin and lincomycin), chloramphenicol, the tetracyclines (e.g., tetracycline, doxycycline, and minocycline) and mupirocin (pseudomonic acid).
Another group of antibacterial agents are antimetabolites that interfere with bacterial synthesis of folic acid. Inhibition of folate synthesis leads to cessation of cell growth and, in some cases, to bacterial cell death. The principal antibacterial antimetabolites are sulfonamides (e.g., sulfisoxazole, sulfadiazine, and sulfamethoxazole) and trimethoprim.
Yet a further group of antibacterial compounds affects nucleic acid synthesis or activity. These agents include the quinolones (e.g., nalidixic acid and its fluorinated derivatives norfloxacin, ciprofloxicin, ofloxacin, and lomofloxacin), which inhibit the activity of the A subunit of DNA gyrase, rifampin, nitrofurantoin, and metronidazole (which not only has activity against the electron transport system but also is believed to cause DNA damage).
BPI protein products are also described to have antibacterial activities in U.S. Pat. Nos. 5,198,541 and 5,523,288 and International Publication No. WO 95/08344 (PCT/US94/11255), all of which are incorporated by reference herein, disclosing activity against gram-negative bacteria, and U.S. Pat. Nos. 5,578,572 and 5,783,561 and International Publication No. WO 95/19180 (PCT/US95/00656), all of which are incorporated by reference herein, disclosing activity against gram-positive bacteria and mycoplasma, and co-owned, co-pending U.S. application Ser. No. 08/626,646, which is in turn a continuation of U.S. application Ser. No. 08/285,803, which is in turn a continuation-in-part of U.S. application Ser. No. 08/031,145 and corresponding International Publication No. WO 94/20129 (PCT/US94/02463), all of which are incorporated by reference herein, disclosing activity against mycobacteria.
BPI protein products have been shown to have additional antimicrobial activities. For example, U.S. Pat. No. 5,646,114 and International Publication No. WO 96/01647 (PCT/US95/08624), all of which are incorporated by reference herein, disclose activity of BPI protein products against protozoa.
Bactericidal/permeability-increasing protein (BPI) is a protein isolated from the granules of mammalian polymorphonuclear leukocytes (PMNs or neutrophils), which are blood cells essential in the defense against invading microorganisms. See Elsbach, 1979, J. Biol. Chem., 254: 11000; Weiss et al., 1987, Blood 69: 652; Gray et al., 1989, J. Biol. Chem. 264: 9505. The amino acid sequence of the entire human BPI protein and the nucleic acid sequence of DNA encoding the protein (SEQ ID NOS: 2 and 3) have been reported in FIG. 1 of Gray et al., J. Biol. Chem., 264:9505 (1989), incorporated herein by reference. Recombinant human BPI holoprotein has also been produced in which valine at position 151 is specified by GTG rather than GTC, residue 185 is glutamic acid (specified by GAG) rather than lysine (specified by AAG) and residue 417 is alanine (specified by GCT) rather than valine (specified by GTT). An N-terminal fragment of human BPI possesses the anti-bacterial efficacy of the naturally-derived 55 kD human BPI holoprotein. (Ooi et al., 1987, J. Bio. Chem. 262: 14891-14894). In contrast to the N-terminal portion, the C-terminal region of the isolated human BPI protein displays only slightly detectable anti-bacterial activity against gram-negative organisms and some endotoxin neutralizing activity. (Ooi et al., 1991, J. Exp. Med. 174: 649). An N-terminal BPI fragment of approximately 23 kD, referred to as rBPI23, has been produced by recombinant means and also retains anti-bacterial including anti-endotoxin activity against gram-negative organisms (Gazzano-Santoro et al., 1992, Infect. Immun. 60: 4754-4761). An N-terminal fragment analog designated rBPI21 has been described in co-owned, co-pending U.S. Pat. No. 5,420,019.
Three separate functional domains within the recombinant 23 kD N-terminal BPI sequence have been discovered (Little et al., 1994, J. Biol. Chem. 269: 1865). These functional domains of BPI designate regions of the amino acid sequence of BPI that contributes to the total biological activity of the protein and were essentially defined by the activities of proteolytic cleavage fragments, overlapping 15-mer peptides and other synthetic peptides. Domain I is defined as the amino acid sequence of BPI comprising from about amino acid 17 to about amino acid 45. Initial peptides based on this domain were moderately active in both the inhibition of LPS-induced LAL activity and in heparin binding assays, and did not exhibit significant bactericidal activity. Domain II is defined as the amino acid sequence of BPI comprising from about amino acid 65 to about amino acid 99. Initial peptides based on this domain exhibited high LPS and heparin binding capacity and exhibited significant antibacterial activity. Domain III is defined as the amino acid sequence of BPI comprising from about amino acid 142 to about amino acid 169. Initial peptides based on this domain exhibited high LPS and heparin binding activity and exhibited surprising antimicrobial activity, including antifungal and antibacterial (including, e.g., anti-gram-positive and anti-gram-negative) activity. The biological activities of peptides derived from or based on these functional domains (i.e., functional domain peptides) may include LPS binding, LPS neutralization, heparin binding, heparin neutralization or antimicrobial activity.
Of interest to the background of the present invention are dye indicators of membrane potential which have been available for many years and have been employed to study cell physiology. These potentiometric dyes are organic compounds whose spectral properties are sensitive to changes in membrane potential. They can be classified generally into xe2x80x9cfastxe2x80x9d dyes, which can follow changes in potential in the millisecond range, and xe2x80x9cslowxe2x80x9d dyes, which generally operate by a potential-dependent partitioning between the extracellular medium and either the membrane or the cytoplasm. This partitioning of slow dyes occurs by redistribution of the dye via interaction of the voltage potential with ionic charge on the dye. Slow dyes include three general chromophore types: cyanines [such as Di-O-C6(3) and Di-S-C2(5)], oxonols [such as oxonol-VI and DiS-BaC2(3)] and rhodamines [such as rhodamine-123 and TMRE JPW-179]. [See Loew, Chapter 8 in Biomembrane Electrochemistry, Blank and Vodyanoy, eds., American Chemical Society, Washington, D.C. (1994), pages 151-173.]
The cyanine class of dyes are symmetrical molecules with delocalized positive charges. Depending on the nature of the dye and its concentration, the potential-dependent uptake can produce either an increase or a decrease in fluorescence intensity. In general, accumulation of the dye and membrane binding leads to enhancement of fluorescence. At high lipid-dye ratios, however, many of the cyanine dyes tend to aggregate, resulting in fluorescence self-quenching. Most carbocyanine dyes with short (C1-C6) alkyl chains stain mitochondria of live cells when used at low concentrations (xcx9c0.5 xcexcM or xcx9c0.1 xcexcg/mL); those with pentyl or hexyl substituents also stain the endoplasmic reticulum when used at higher concentrations (xcx9c5-50 xcexcM or xcx9c1-10 xcexcg/mL). The cyanine dye DiOC6(3) (3,3xe2x80x2-dihexyloxacarbocyanine iodide) has less tendency to aggregate and displays an increased fluorescence quantum yield as it binds to the subcellular membranes. DiOC6(3) is lipophilic and is often used as a stain for mitochondria and endoplasmic reticulum in eukaryotic cells.
The green fluorescent cyanine dye JC-1 (5,5xe2x80x2,6,6xe2x80x2-tetrachloro-1,1xe2x80x2,3,3xe2x80x2-tetraethylbenzimidazolylcarbocyanine halide, available as an iodide from Molecular Probes or as a chloride from Biotium, Inc.) exists as a monomer at low concentrations or at low membrane potential. However, at higher concentrations (aqueous solutions above 0.1 xcexcM) or at higher potentials, JC-1 forms red fluorescent xe2x80x9cJ-aggregates xe2x80x9d that exhibit a broad excitation spectrum of 485 to 585 nm and an emission maximum at xcx9c590 nm. Emission from this dye has been used to investigate mitochondrial potentials in live cells by ratiometric techniques. Various types of ratio measurements are possible by combining signals from the monomer (absorption/emission maxima xcx9c510/527 nm in water) and the J-aggregate. Optical filters designed for fluorescein and tetramethylrhodamine can be used to separately visualize the monomer and J-aggregate forms, respectively, or both forms can be observed simultaneously using a standard fluorescein longpass optical filter set.
The oxonols are anionic molecules that also show enhanced fluorescence upon binding to membranes. However, because of their negative charge, binding of oxonols is promoted by depolarization of the membrane. The negative charge of oxonols also lessens intracellular uptake and reduces their association with intracellular organelles.
Rhodamine-123 is a cell-permeant, cationic, fluorescent dye that is readily sequestered by active mitochondria without inducing cytotoxic effects. Uptake and equilibration of rhodamine-123 is rapid (a few minutes) compared to dyes such as DASPMI, which may take 30 minutes or longer, and this dye is especially suited for flow cytometry applications. Mitochondria stained with the dye appear yellow-green through a fluorescein longpass optical filter and red through a tetramethylrhodamine longpass optical filter. Unlike the lipophilic rhodamine and carbocyanine dyes, rhodamine 123 does not stain the endoplasmic reticulum. Rhodamine-123 has been used with a variety of cell types including nerve cells, bacteria, plant cells and spermatozoa, and has also been used to study apoptosis, axoplasmic transport of mitochondria, bacterial viability and vitality, mitochondrial enzymatic activities, transmembrane potential and other membrane activities, multidrug resistance, mycobacterial drug susceptibility and oocyte maturation. Derivatives of rhodamine-123 such as TMRE have been developed that are more permeable and have less hydrogen-bonding interaction with anionic sites in the mitochondrial inner membrane and matrix.
There continues to exist a need for novel antimicrobial agents useful for treating a variety of infections and for methods of identifying such novel compounds. Such methods ideally would provide for rapid and highly selective identification of compounds that may be structurally distinct from the major conventional antimicrobial agents.
The present invention generally provides methods for identifying antimicrobial compounds (including, for example, antifungal or antibacterial compounds) based on the discovery that a class of antimicrobial agents based on or derived from bactericidal/permeability-increasing protein (BPI) generates unique effects on fungal and bacterial cells as revealed by treatment with a cyanine membrane potential indicator dye, DiOC6(3). When BPI-derived peptide compounds are employed as antifungal agents, their effects are characterized by localization of the cyanine dye to mitochondria with increasing accumulation of this dye in a peptide concentration-dependent manner, and with retention of the dye notwithstanding an onset of loss or reduction of fungal cell viability at the same peptide concentration. The dye appears to be retained even after cell death has occurred (as confirmed, e.g., by a negative 24-hour growth culture or by use of other viability indicators, such as propidium iodide). Similarly, when rBPI21 and BPI-derived peptide compounds are employed as antibacterial agents, their effects are also characterized by increasing accumulation of this cyanine dye in a peptide concentration-dependent manner, with retention of the dye notwithstanding an onset of loss or reduction of bacterial viability at the same peptide concentration.
Novel antimicrobial agents may be rapidly and selectively identified by screening test compounds for replication of the characteristic increase in dye fluorescence intensity produced by BPI protein products with continued retention of dye notwithstanding loss (i.e., reduction) of viability within the tested target cell population. Sources of test compounds include, for example, libraries (including combinatorial libraries) of inorganic and organic compounds (for example, bacterial, fungal, mammalian, insect or plant products, peptides, peptidomimetics and organomimetics). Presently preferred standard BPI-derived antimicrobial peptides that are known to produce this characteristic pattern include XMP.391 (SEQ ID NO: 1) and XMP.445 (SEQ ID NO:2). This aspect of the invention thus contemplates a method of identifying an antimicrobial agent, particularly an antifungal compound, comprising the steps of: (a) contacting a target cell (e.g., a fungal cell or a bacterial cell) with a test compound and with a membrane potential indicator dye, and (b) detecting an increasing accumulation of this dye and retention of this dye despite loss or reduction of target cell viability. Presently preferred membrane potential indicator dyes are DiOC6(3), JC-1, rhodamine-123 and MitoTracker Red CM-H2Xros [M-7513, Molecular Probes, Inc., Eugene, Oreg.]. The concurrent loss or reduction of target cell viability may be confirmed by routine culture or through use of viability dyes, such as propidium iodide.
It is further contemplated that screening methods according to the present invention may involve multiple further stages of screening, including selection of test compounds that have a differential effect on target cells in comparison to non-target cells (e.g., a reduced effect on mammalian cells relative to fungal cells). This aspect of the invention provides a further screening assay involving (a) contacting a mammalian cell with the test compound and with the membrane potential indicator dye, and (b) detecting no substantial increase in dye fluorescence intensity. Optionally, compounds may be screened for selectivity for one type of microbial cell, e.g., selectivity for bacterial vs. fungal cells or vice versa.
Test compounds may be alternatively or additionally assayed for ability to kill or inhibit growth of target cells (e.g., fungal cells or bacteria) in vitro using any assays known in the art, including broth or radial diffusion assays. Suitable compounds may have a 2-fold, 10-fold, 50-fold, 100-fold, or greater separation (selectivity) between antimicrobial activity and mammalian cell activity. The most desirable compounds will preferably have a 50-fold or greater separation between antimicrobial activity and mammalian cell activity as quantified by differential effects on dye fluorescence intensity.
The in vivo antimicrobial activity of test compounds may also be assayed in any animal models of infection known to those skilled in the art. Such assays include those for in vitro and in vivo oral availability and those for in vivo oral activity as evidenced by activity when administered orally in a comparative survival study.
Another aspect of the invention provides kits for use in conducting the screening methods of the present invention. Such kits may optionally include (a) a membrane potential indicator and (b) a BPI-derived antimicrobial peptide or other BPI protein product suitable for use as a standard (positive control) against which the test compound may be compared.
The present invention also provides novel antimicrobial compounds identified by the screening methods of the present invention.
Yet a further aspect of the invention contemplates the treatment of infections, including fungal and bacterial infections, using compounds identified by the screening methods of the present invention that exhibit the above-described characteristic pattern, other than compounds known in the art (including BPI protein products such as BPI-derived peptides).
Numerous additional aspects and advantages of the invention will become apparent to those skilled in the art upon consideration of the following detailed description of the invention which describes presently prepared embodiments thereof